In this talk, we propose collaborative algorithms for wireless networks. Assuming that nodes communicate using frequency division multiple access, we propose collaborative bandwidth allocation strategies for the above clustering protocol that minimize proportional blocking probabilities. The strategies utilize a very elegant form of utility function, which allows for easy optimization. We then subsequently propose optimal rate allocation among the nodes based on joint encoding of the data from these nodes assuming that they are correlated. The correlation model assumed is Gaussian. Encoding rates are from a relaxed convex rate distortion region using a special encoding procedure. We develop an algorithm for allocating rates subject to energy and buffer constraints. Subsequently, we incorporate joint rate and power control algorithms into the nodes to cater to high data rate communications. This way, we address the fundamental tradeoff that exists between power levels, data rates, and congestion rates in the network. Lastly, we assume that nodes communicate using code division multiple access and we propose a robust receiver for uplink communication. The receiver uses low order auto-regressive models to approximate the multi-path fading channel taps, and a post correlation-based uncertain model for estimation purposes.
Hence in totality, we will throw some light from information theory to collaborative signal processing to help design optimally functional wireless networks.